Design and Implementation of Literature Information Mining and Analysis Tools Based on Large Language Models

With the rapid growth of academic publications, researchers are facing a severe problem of information overload. Traditional intelligent literature analysis methods often rely on rule-based text mining or conventional machine learning techniques, which typically achieve only keyword retrieval or shallow content analysis, making it difficult to meet the need for in-depth understanding of academic literature. With the fast development of large language models (LLMs) in natural language processing and their increasing applicability to such research tasks, this study designs and implements an intelligent literature analysis tool based on LLMs to enhance both the efficiency and depth of researchers’ reading and comprehension processes. The system adopts Zhipu AI’s GLM-4 model as the core analytical engine, and builds a web platform with the Flask framework. It supports PDF uploading, text parsing, key information extraction (research questions, methods, and findings), structured abstract generation, and interactive question answering. Experimental results demonstrate that the tool outperforms traditional methods in terms of accuracy and efficiency of information extraction, while also offering good user experience and practical value. At the same time, this study contributes to the academic community by verifying both the feasibility and limitations of LLMs in the context of scholarly literature analysis, proposing an integrated processing framework, and revealing through experiments on both Chinese and English texts the performance differences in cross-lingual processing. These findings provide useful insights for further improvement and extension.